Model comparison

AMD-OLMo-1B-SFT-DPO vs Qwen3-1.7B

At Q4_K_M, AMD-OLMo-1B-SFT-DPO is the smaller download — 733,520,640 bytes against 1,107,409,472.

From the file· summed bytes, KV per layer

Side by side

AMD-OLMo-1B-SFT-DPOQwen3-1.7B
Parameters1.2B2.0B
Architectureolmoqwen3
Layers1628
Native context2,04840,960
Mixture of expertsnono
Quantizations published1848
Smallest quantization0.53 GiB0.50 GiB
Q4_K_M0.68 GiB1.03 GiB
Licenceapache-2.0

KV cache by context

the term that decides long-context viability
ContextAMD-OLMo-1B-SFT-DPOQwen3-1.7BRatio
4,0960.50 GiB0.44 GiB1.14×
8,1921.00 GiB0.88 GiB1.14×
16,3842.00 GiB1.75 GiB1.14×
32,7684.00 GiB3.50 GiB1.14×
65,5368.00 GiB7.00 GiB1.14×
131,07216.00 GiB14.00 GiB1.14×